What does P-Value mean in Regression? - YouTube

Ever had difficulty in understanding what exactly is p-value.

this tutorial gives you

a brief idea on how you can use p-values in your hypothesis testing.

One fine day when

I was working on a regression problem, I came across such a screen and got a bit confused

as to which features a significant based on the p-values which are denoted on the rightmost

column, so I decided to make this video so that none of you are ever confused because

of p-value.

Let's understand first what a NULL hypothesis is.

NULL hypothesis is basically

states that nothing new is happening and the old theory is true for regression it simply

means that there is no relation between the dependent and independent variable and the

coefficients of independent variable denoting the dependent variable are zero.

Now what is

p value mean p value is basically used in hypothesis testing to help you support or

reject the null hypothesis. p-value is an evidence against your null hypothesis smaller the value

stronger the evidence to reject the null hypothesis.

A simple example to give you an idea of how

null hypothesis and p value work is say, for example a pizza place claims that their

delivery times are 30 minutes or less on an average but I think it's more than that so

what do I do?I conduct a hypothesis test.

My null hypothesis is that the mean delivery

time is 30 minutes and my alternate hypothesis which I want it to be true is that the mean

time is greater than 30 minutes. I randomly sample some delivery times and run the data

through the hypothesis test and my p value turns out to be 0.001.

In simple terms that

there is a probability of 0.001 that you will mistakenly reject the pizza places claim

that they're delivery time is less than or equal to 30 minutes since typically we are

willing to reject the null hypothesis when this probability is less than 0.05 you conclude

that the pizza placed is actually wrong.

So if your p value is less than 0.05 reject the

null hypothesis if it's greater than 0.05 do not reject the null hypothesis.

Now coming

back to our problem. I have five variables age, time, last contact age, total call count

success call counter and all of this and I have a p value associated with each variable which

are my significant variables in which are not my significant variables is what my p

value is going to signify.The p value for each term test for the null hypothesis that

the coefficient is equal to 0 no effect so a low p value which is less than 0.05 indicates that you

can reject the null hypothesis in other words a predator that has a low p value is likely

to be a meaningful addition to your model because the changes in the predictor's value are related to the changes in response

variable. Conversely a larger p value suggest that changes in the predictor are not associated

with the changes in the response.

So in our case the p value for the first three variables

are less than 0.05 so I reject the null hypothesis meaning that the weights or the coefficients

are not 0 and except the alternate hypothesis, which is their standard coefficients which

are given after the regression has been computed and in simple terms the first three terms

having p values less than 0.05 contribute to the predictor that is my final output variable.

Thank you for watching this video. I hope that you found this video informative. Do subscribe

to the channel on the link given below. Thank you.